Remember when iTunes first came out, and nobody could figure out how to get their music from their computer to their iPod?
When Apple released iTunes, the desktop music streaming and file operations worked great but nobody could figure out how to put music on their iPod. This could be because Apple neglected usability testing.
When done properly, usability testing can reveal the shortcomings in a product’s performance by mimicking the ways that consumers will interact with the product post-release.
However, if you don’t perform usability tests correctly, you could end up with misinformation that instead makes your product worse.
Here are some of the more common usability mistakes, as identified by Codal’s UX specialists.
1. Not Videotaping the Tests
Videotaping usability tests not only allows you to improve test proctoring skills, but also provides a level of detail that can’t be reached with text based results alone.
When you record the usability test in writing as opposed to on tape, you lose the subtleties of interaction. On video recordings, you could watch the user lean toward the screen in confusion while they search for their objective, for example.
As you become more familiar with usability testing, you can start to pick up the rhythm of the interaction between the user and the app or website. You may notice how the user becomes impatient and starts blindly clicking every link in sight (if your UX is poor).
When the test is over, having a video file to show the rest of the UX team will help them to remember what issues were had by users.
The best camera angle for usability testing places the camera in a position where it can see the monitor and the user simultaneously. Typically the proctor can stand a few steps directly behind the seated user for optimal footage. If possible, use multiple cameras.
Remember: get permission from the tester before filming them.
If they don’t give permission, or you don’t have a camcorder on hand, you can always download a screen recorder app such as IceCream.
2. Poor Setup
It’s easy to underestimate the amount of preparation involved in usability testing. Before the first set of tests are performed, it’s absolutely critical to perform multiple pilot tests. Not only will these pilot tests iron out any potential technical issues, but the UX specialist proctoring the test will benefit from the extra practice rounds.
When it comes to choosing the appropriate number of people for your test, five appears to be the ideal number (Alex Baker).
Make sure you try to address all of the hidden variables from test to test. For those of you who don’t have your middle school notebook handy, the Scientific Method invokes that we only change one variable each round of testing.
If a rogue user decides he wants to open the site in Opera rather than the more standard Google Chrome or Internet Explorer, it could result in a statistical outlier.
Maybe another time the user had 17 other windows open in the background, impairing focus and available RAM. This relates also to the important concept of controlling the test environment. Chances are, if your tester managed to open that many windows prior to being tested, he was not being monitored closely enough.
Take time to ensure that the only thing that changes between each round of testing is the person sitting in front of the screen and the hands on the clock.
3. Not Asking the Right Questions
You must be very particular about how you phrase your tasks to the user.
According to Jared Spool of uie.com, furniture retailer Ikea once redesigned their website based on poor usability testing.
During the test, users were instructed to simply “Find a Bookcase”. Those instructions may be clear and concise according to an English Teacher, but they don’t serve the purpose of the Usability Test well.
As one might expect, each and every participant in the usability test went to the nearest search engine and entered “bookcase”.
Because Ikea was so concise with their phrasing, usability test returned 200 examples of the same action.
Upon consulting UIE Magazine, Ikea changed the phrasing to “You have 200+ books in your fiction collection, currently in boxes strewn around your living room. Find a way to organize them”.
This new phrasing doesn’t blatantly give away the desired action. Users now have to decide what type of furniture they want to store their vast collection of fiction in.
During the second round of usability testing under the new phrasing, not a single test subject entered “bookcase”. Instead, the test subjects entered the search queries “shelves” and “storage systems”.
Most of the users by-passed the search bar completely and instead clicked through the categories of the site, presumably looking for the most aesthetically pleasing solution.
When designing the task prompts for a usability test, make sure you don’t give away the solution. You want to force the test subject to make realistic decisions.
4. Choosing the Wrong Participants: Don't Do it!
Unless you’re in the banana retail industry, you wouldn’t want to enlist a squad of primates to test your product.
“This website needs more bananas, ASAP!” (RWBY Images)
In fact, even if you were in the banana retail industry, you still wouldn’t want primates to be testing your bananas. Because they have far more experience picking fruit, and have been exposed to many different qualities, primates wouldn’t replicate typical human behavior.
Likewise, it wouldn’t make sense to enlist a panel of software designers for the purpose of usability testing. It depends on the product you’re testing, but much like the primates, software designers won’t accurately represent the way that a typical consumer will navigate through a site.
5. Starting with a Poor Research Question
Many people start usability tests with the goal of improving a given website or product. While the underlying goal of usability testing is product improvement, for the purpose of designing a test, it’s better to use the “Search and Destroy” approach:
That is, narrow your research question prior to assigning a task to your testers:
1. Find problems or inefficiencies with the current design. [Search]2. Fix those problems. [Destroy]3. Rinse and repeat.
The strength of a usability test lies in the fact that it allows you to identify the elements of a website that are confusing the user.
Rather than “will our customers get to the checkout section quickly?”, a good research question for usability testing would be “will there be a fewer calls to our tech support team now that we’ve switched “item a” and “item b”?”
Conclusion: If You Decide to UI Test, Do It Right!
If you don’t execute a usability test properly, the information that you get could cause you to make changes to your website based on false information.
However, when done properly, usability testing offers the opportunity to learn about both the target market and the inefficiencies of the product in its current build.
If you’re designing a program that is intended to handle music files, you better make sure that customers can easily figure out how to transfer the music files to their portable device.
Regardless of what you are testing, you may find that working with an expert UX design agency will give you the edge in identifying these inefficiencies and usability problems.
